A TISM-Bayesian network hybrid approach for assessing supply chain resilience

Abstract

Supply chains operate in a global market that has significantly increased their complexity and exposure to disruptions. Erratic changes in product demand and consumption further intensify the challenges of planning and designing a resilient supply chain. To succeed in such a dynamic environment, it is imperative to estimate supply chain resilience effectively. This study investigates the enablers that influence supply chain resilience through an exhaustive analysis of the literature and expert opinions. Fourteen key enablers were identified and shortlisted for further study. To enhance resilience capabilities, it is essential to identify, interpret, and acknowledge the critical proficiencies of firms, which serve as the foundation for developing an appropriate model. Consequently, this research constructs a resilience analysis model by integrating Total Interpretive Structural Modeling (TISM) with Bayesian Networks (BN). TISM is employed to analyze the interdependencies among the enablers by leveraging the knowledge and experience of domain experts. This process results in a hierarchical structure comprising ten levels, which illustrate the driving power and interdependencies among the enablers. Enablers such as Supply Chain Network, Collaboration and Coordination, and Information Sharing exhibit the highest driving power, influencing other enablers significantly. Conversely, Revenue Sharing emerges as the most dependent enabler, relying heavily on others. The proposed TISM model is subsequently transformed into a Bayesian Network, where expert opinions are converted into conditional probability sets. This probabilistic approach helps identify critical indicators among the enablers, which are essential for assessing supply chain resilience. To demonstrate the applicability of the proposed methodology, two case studies were conducted in the automotive and steel sectors. The results highlight the critical indicators in each case that require focused attention to enhance supply chain resilience. These findings offer actionable insights and theoretical support for enterprises in the automotive and steel industries, enabling them to adapt to the complexities of the modern business environment.